The paper reviews the developments in the short-term forecasting of rainfall and flow using weather radars, especially with regard to their use for flood warning. The methods of radar rainfall forecasting considered range from simple field advection and feature tracking approaches through numerical weather prediction models to simple cloud models incorporating the assimilation of weather radar, satellite and weather station observations. Tailoring of catchment runoff models to make better use of weather radar data has led to the development of grid models whose parameterisations are supported by digital terrain models. An assessment of this distributed approach to flood forecasting is presented based on results from the new IH Grid Model. A recent work at the Institute of Hydrology has resulted into an advanced River Flow Forecasting System ( RFFS ) Hydrological Kernel, a software product for operational forecasting and off-line model calibration. It is also a purpose of this paper to describe its implementation in Yorkshire and to briefly report on other applications and developments of the RFFS in the UK and overseas. These include the recent integration of a weather radar processing and display system, called Hydrological RADdar system (HYRAD), into the RFFS forecasting environment.